Dechallenge is what happens to an adverse event when the suspected drug is stopped or reduced; rechallenge is what happens when the drug is given again. A positive dechallenge means the event resolves or improves after withdrawal. A positive rechallenge means it returns on re-exposure, strong evidence that the drug caused it; a negative rechallenge means it does not return. Rechallenge is often not done, and the WHO-UMC guidance notes that a rechallenge is rarely ethically justified. Both causality methods weigh these observations. In WHO-UMC, the response to withdrawal must be plausible for 'certain' and clinically reasonable for 'probable/likely', and a satisfactory rechallenge is asked for only under 'certain', where necessary. The Kramer algorithm scores dechallenge and rechallenge as two of its six axes. Burna AI applies both methods to each drug in a regimen and lists the WHO-UMC criteria each case meets.
More on multi drug attribution.
- What is multi-drug attribution?
- How does Burna AI compute adverse event attribution?
- What are the WHO-UMC causality categories and the Kramer algorithm?
- Which oncology regimens does Burna AI cover?
- Can Burna AI grade single-agent trials too?
- How does Burna AI handle protocols where the trial drug profile is sensitive sponsor IP?